Degree Name

Master of Science - Research


Centre for Medical Radiation Physics


Current and emerging photon radiation therapy modalities require dosimetry with tissue-equivalence, high spatial resolution, the possibility to measure surface dose and dose inside the target, and preferably the ability to use the system in-vivo and online. In fact, the high dose gradients a fforded in both intensity modulated radiation therapy (IMRT) and microbeam radiation therapy (MRT) require the small size of solid-state dosimeters for the accurate quantification of absorbed dose to tissue both pre-treatment and for real-time quality assurance purposes. However, with much of the advances made in solid-state dosimetry having focused on the use of silicon as the base sensing medium, an over-response at low x-ray energies (< 150keV) is observed due to the increased chance of photoelectric absorption as compared to water. To remove this dependence, and thus improve dosimetric accuracy, a matching of atomic composition is required. Carbon based radiation detectors provide an exciting opportunity to help fulfill this need, and as such, form the basis of study in this work. In particular, the research carried out aims to investigate the online x-ray irradiation response of three novel detectors made from either synthetic single crystal diamond or carbon nanotubes (CNTs). Indeed, the latter display extraordinary electronic and optical properties; and by investigating the dosimetric properties of each type of detector in this way, their potential as clinical medical radiation dosimeters can be evaluated.

The first detector investigated was a novel synthetic single crystal diamond detector implanted with 2MeV boron ions at the depth of the Bragg peak. Due to radiation damage which causes a complete graphitisation of the diamond crystal at that depth, a tissue-equivalent contact is formed in addition to a small sensitive volume ~ 1.5 µm thick; this makes the detector ideal for accurately measuring the micrometer sized dose profile present in MRT. Preliminary testing under orthovoltage and linac generated x-ray beams demonstrate excellent stability, repeatability, dose-rate linearity, and a relatively flat energy response that matches with theoretical calculations. In addition, the need for a priming dose which is common to most diamond detectors was removed due to the high purity of diamond used. These dosimetric properties are very promising and would suggest the detector design is suitable for further testing under intensive synchrotron generated x-rays at the European Synchrotron Radiation Facility in France. However, a dose sensitivity upwards of 68 times greater than what the theoretically induced photocurrent will allow was also observed, suggesting the bias potentials applied (minimum of - 2.5 V) contribute an electric field strength which is able to penetrate the insulating diamond substrate well beyond the depth of the buried boron electrode. This increase in the thickness of the sensitive volume renders the detector inoperable for use as an edge-on dosimeter in MRT, an application for which it was specifically designed, although its application in IMRT is still possible but which requires further study.

The second detector tested was a U.S. patented device made of a thin percolation network of single-walled CNTs. Commercially known as FlexDosTM, it is claimed by the manufacturers ARTsensing Inc. that the detector is suitable for use as a real-time in-vivo measure of surface dose due to the induced photoresponse obtained. However, dosimetric characterisation of the device was hampered by the discovery under scanning electron microscope that the thin CNT layer did not exist in any of the four devices purchased, with an observed signal produced only a result of air ionisation at the contacts. Whether the CNTs were never deposited on the supporting substrate or were accidently removed remains unclear, although the latter would indicate the device is not rugged enough for use in the clinical environment. Since neither the commercial manufacturer nor the inventors replied to queries regarding the fact that no CNT layer existed, it remains unconfirmed whether the thin CNT film produces a genuine signal under irradiation. What is clear however is that the design needs further enhancement to remove the collection of charge from air ionisation at each electrode interface.

The third and last detector tested was a free-standing thin single-walled CNT film known as buckypaper. A thicker variant of the FlexDos design, irradiation under a 6MV x-ray beam showed no evidence of an online response.

In summary, this study indicates that boron implanted single crystal diamond detectors have a great potential for improving dosimetric accuracy in IMRT as high resolution real-time in-vivo dosimeters, although their application in MRT requires a refined detector design. Conversely, randomly deposited thin CNT films appear unable to measure an online photoresponse, although the effects at the percolation level remain inconclusive.